+1 vote
How many years do you go out with your Annual Subscription LTV calculations? Assuming that you launched annual subscriptions less than 12 months ago and have no annual retention data yet, therefore you need to make assumptions on annual renewals, how many years does your model include?
by (240 points)
I'm a bit confused by this question. Are you asking about how to predict your 12 month LTV when you have < 12 months of data?
Let's say you just launched an annual subscription option, therefore you have no renewal rates yet, your 1 year LTV would simply be the annual subscription price, but what about your 2 year? Lets say you assume a 50% renewal rate, you would then have a 2 year LTV of (1.5 * annual subscription price), what about the 2nd year renewal? I'm asking how many years of assumptions do most people use? You can hypothetically go out 4-5 years if you wanted.
so your question is more about predicting LTV? how far out did you want to predict to?
Thats my question, I want to see what is common in the industry.

2 Answers

+2 votes
Working on a mobile product with a subscription-based model.

For our 6m auto-renewable subscription, we're taking only 1 renewal after the first payment. So, LTV is calculated for 6m + 1 day. For shorter subscriptions, LTV is calculated for 6m. Anything earned beyond that period will be an additional plus.

Pretty conservative approach I would say, but allows us to be on the safer side and more rigorously control our CAC, improve LTV and ensure LTV > CAC.

Worth to mention that it really depends on your business targets and limitations, product peculiarities and the niche you're working in.
by (350 points)
+2 votes

I'm working on a project for a subscription app that only has a 1-year option, and we are buying against the 1-year LTV. The logic behind this is pretty straightforward:

  1. We don't have any data on cohorts that are older than one year, and we don't want to make any big assumptions around renewal rates absent any data;
  2. When you start thinking about multi-year LTV, you need to think about discounting, which in general overcomplicates things. See: Should Lifetime Customer Value (LTV) be Discounted?;
  3. In general I'm a fan of bidding up conservatively and iteratively expanding the frontier of my LTV projection. See: It’s time to retire the LTV metric and Optimizing campaign spend with the Blended Same-Month Return metric;
  4. The market can change dramatically in a year -- do we really want to take large financial bets on discrete actions (renewing a purchase) that customers will make in a year's time? That feels like irresponsible risk to assume.

In general, I think this is a great question, and teams don't spend enough time thinking about it. An imprudent decision I see companies make is extending out their payback window for long-term subscriptions after raising a huge round of financing because they need to hit aggressive growth targets. A better way to manage that necessity, in my opinion, is to expand the product catalogue and create new opportunities to monetize that can drive top-line growth in ways that aren't tied to a one year cadence.

by (7.7k points)